With Instagram becoming more popular, a lot can be done with images and videos, but it is equally expensive to process or resize them. The traditional way of editing the videos and images would be to host a VM that is always running and when an image or video arrives for editing then running the application within the VM. But we have to incur costs for running the VM 24/7 without it being used all the time. Serverless could be the best option for these types of jobs. Serverless services can be used to dynamically resize the image or change the video transcoding for different target devices.
Also, with serverless, we can tap into the Google Vision API or Amazon Recognition to recognize faces and images or flat inappropriate content. In the AWS chapter, we created a tutorial where an image is uploaded into one S3 bucket and then Lambda is automatically triggered when the image is uploaded to the S3 bucket to resize it and push it to another bucket.